Linux on the (consumer) Desktop
people.gnome.org
people.gnome.org
It's probably just about the opposite. Linux is many years away from the level of plug-it-in-and-forget-it integration that Windows has. Linux is light years away from having anything resembling Active Directory in quality. And like it or not, the corporate world lives and dies by Office.
I'd say if you want to make progress, the consumer space is the only possible bet, and it won't be on desktop computers but on tablets and embedded devices which, for the most part, will replace people's keyboard and monitor machines at home.
But not in the office.
There is also support, training, excellent literature and documentation, large quantities of staff, functional software at a reasonable price and as you mention: Active directory (which is awesome if you actually understand it).
If you've seen some of the shit I've seen wired up to Office in the last few years, you'll get it. It's just awesome. Entire businesses hang of it. It might not be clean and squeaky but it's functional and does the job. Google Docs + iWork + LibreOffice are precisely zero competition.
(I have to add this to each of these posts to avoid the shill downvotes: I am primarily a Linux user which suits me but I don't think it'd suit most people).
>the consumer space is the only possible bet
Below is a link to a discussion we had here 2 weeks ago about Ubuntu at Google. Although engineers at Google can choose a different one, Ubuntu is the desktop OS with the most support from Google's "IT department". In contrast, Windows is the only choice that requires some sort of approval from management.
Also, although it is a paying customer for Canonical's Ubuntu Advantage support program, Google is not even Canonical's largest business desktop customer.
As long as we do not have unrealistic aspirations of completely replacing Windows in the enterprise, why not try to expand on the presence desktop Linux already has there?
I'm assuming you've setup OpenLDAP and one of open Kerberos implementations (e.g. MIT) and have found them lacking in some way compared with AD. If so, could you expand on where they were lacking in quality? I have used both and find open alternatives much easier to extend, and not any harder to setup.
Here's an interesting read that argues the opposite of your point: ftp://ftp.uni-duisburg.de/LDAP/Adam-Eval1-0.pdf
Office, Outlook, most AAA games, itunes. Don't tell me that because you can run them today on your machine (well maybe) that that somehow translates to what common users have experienced over the last decade.
Outlook is a rather heavy monstrosity. There are better e-mail clients out there, including those on Linux.
As for Outlook - while I've set up my home computer to not use it, for most it's just easier to use it. Convenience has a lot going for it. For non-tech-savvy people it's also a lot easier, since then calls to tech support at work will actually get you help. :)
I use Linux machines for what they do well, and pretty much all my machines sans the "main" Laptop are Linux-driven.
That said: This statement cannot be refuted enough.
Libre Office and its forks are merely bad imitations of previous versions of MS Office while offering few to no benefits once you exclude cost. They are not "decent" alternatives. They are free, but noticeably sub-par alternatives. They are in no way drop-in replacements.
If I want to take the cost of working with a low-end Office-suite which covers the bare basics and sticks to a late 90s design-paradigm, I'll go with Google's online-offering because it actually provides some benefits to weigh up for the cost.
And sometimes I do just that. But you'll never find me meddling with the desktop Linux office-suites. They're simply not worth it.
Just basic things like creating sections, chapters, tables, footers, headers and making it all look good and consistent, quickly and easily.
MS Office, since the move to the ribbon-system have really excelled at exposing the functionality required to do this. It's always there in your face, just one click away, when you need it and then it gets out of your way. It's fantastic.
Libre Office and its forks are still stuck in the last generation of UIs where you have a million toolbars, all just mediocrely capable and insufficient to your full needs, none adapted to your current context and always there taking up space and cluttering the UI wheter you need them or not.
There's just no comparison between the user experiences here. MS Office is in a league of its own.
But the reason Microsoft moved from the safe (familiar) UI to the new one was because they realized the old UI-model didn't scale well with expanded functionality.
They decided to try to improve the user experience by leaving behind what clearly did not work.
For a company which owes most of its legacy to users preferring familiarity, that's very much not a small thing. But they decided it had to be done.
More than five years later, Libre Office still haven't done anything at all to fundamentally improve UX. To me that sounds like somebody simply not caring.
For a very long time for most things casual users did Windows was a better fit then linux. Want to play games? Windows by a mile. Burn DVDs? Windows. Download music legally? Windows. Download music illegally? Windows. Word Processing? Windows. Spreadsheets? Windows. Use that new camera/scanner/printer/ipod? Windows by a mile. Read your office email? Windows. DTP? Windows. Visit flash websites? Windows.
People don't want to make a choice or have to care about what they buy. The ISVs are quite happy to ship windows as they have reusable experience and it's just plugging things together.
The reason no market share destroying competition has appeared is that people are quite happy apart from a minority of users who are quite vocal.
consider that are more windows power users than there are mac users and Linux users combined.
I'm a Linux user (heavy with 15 years' experience) but i don't have a problem with the current situation.
Same as Linux in this day and age (if user has such option).
The reason no market share destroying competition has appeared is that people are quite happy apart from a minority of users who are quite vocal.
With this logic any kind of monopoly can be justified, but it doesn't make monopoly good per se. Artificial lack of competition is bad.
consider that are more windows power users than there are mac users and Linux users combined.
Network effect is part of the problem, but the bundling issue is just a method to feed it. So one way to break it - is to remove the bundling blocker to allow fair competition.
I don't get these arguments: the user has the option. In the UK, there are magazines on every shelf in every newsagent pushing Linux and in every PC related stored they are pretty much giving Ubuntu away (yes they even had a box at the checkout in PC World with free CDs - completely untouched). You can buy PC's with it preinstalled and buy PCs with no operating system so you can install it.
I know MANY "normal" non enthusiast and non power users from social housing unemployed to professionals who have used Linux distributions and just end up back with Windows. Everyone knows what Linux is or has heard of it.
I don't think Microsoft is a monopoly - they just work pretty well on the desktop for most users (well they did until Windows 8 but I'm not going there in this discussion ;-). The competition is there but it's not really competition in mindshare as people are comfortable.
If you forcibly removed the bundling, you are going to get nowhere i.e. people saying "Can I just have it with Windows on it". Also, at the sales desk, the salesperson is going to make a few extra readies for selling Windows with a PC so you're going to lose anyway.
Take a look at where IBM is now compared to the late 1980's. They monopolised and beat their competition to death and they didn't survive in that market. Microsoft's time will come if they fuck up or piss off a lot of people, but at the moment they do what people want for a price that people want to pay.
The idealisms of "Microsoft sucks" and is a "monopoly" remind me of Slashdot circa 2001 onwards where OSX and Linux were going to take over the world and destroy Microsoft.
Sorry it ain't going to happen and I'm not really that bothered myself.
Is this the same UK I live in? The newsagents I see might have one Linux magazine lurking behind the iPad magazines. And you really have to go looking for computers that come with Linux preinstalled or no OS. Dell stopped offering them online, and I haven't seen a Linux PC in a retail store since the brief wave of Linux netbooks. I found a small business that sold me a computer with my choice of distro on, but they didn't really promote the option at all.
> If you forcibly removed the bundling, you are going to get nowhere i.e. people saying "Can I just have it with Windows on it"
"Sure, that will be £20 extra" (a complete guess at the OEM Windows price)
Obviously preinstalled Linux systems aren't going to topple Windows overnight - there's a lot of network effects and familiarity there. But at the same time, it would definitely benefit Linux adoption if consumers regularly saw an option to use Ubuntu, say, with known-good hardware and without having to install it themselves. Not many would take it at first, but a few people would go home and look it up.
> Everyone knows what Linux is or has heard of it.
Many people have heard of it, certainly, but they often have some vague idea that it's a command-line interface for ultra nerdy experts. That's why Android and Ubuntu both avoid using the word 'Linux' in their public branding.
That actually is a lack of gratitude and failure to improve Linux image on their part. But I agree with the rest of the above. Removing the bundling won't magically solve all issues, but it's a huge blocker which prevents any kind of solutions from reaching the broader audience.
I don't see it quite like that. Firstly, I'm not sure that the people working on Linux (the kernel) particularly want to be a popular brand - it would mean a deluge of naive questions, apart from anything else.
Secondly, Ubuntu benefits from Linux, GNU tools, X, Gnome, Debian, and many other projects. In techie circles, they readily acknowledge that it wouldn't be possible without the work done in those upstream projects. But trying to fit them into the public branding would confuse users.
This sounds like a fairy tale, given how many problems people have with claiming the Windows refund alone, let alone getting alternative OSes preinstalled. And no OS option? Virtually non existent, except may be for corporate customers and servers.
> I don't think Microsoft is a monopoly
They formally avoid antitrust regulations by delegating the refund to OEMs (see http://en.wikipedia.org/wiki/Windows_refund), but it doesn't change the factual situation - Windows is bundled practically by all major OEMs, and you hardly can find an option of Linux preisntalled or no OS for those OEMs. Getting a refund for Windows is a burden which one has to fight through, do you think it's not an artificial restriction on the competition? I think it is, and it needs to be broken.
I won't go much into the issue other than say that I find it somewhat sad that we are essentially hooking up these children and ultimately our future on the products of a private entity. Even if they are the better option out there, there is still an issue of unnecessarily power being granted to a private entity.
You might also want to checkout what kind of deals your local government/municipality made with MS.
Two things are real blockers for Linux desktop adoption: the Adobe suite (or alt) for professionals, and gaming for casual users. One of these is about to be solved :)
Unfortunately I cannot seem to find any details on the what's backing this distro. Maybe that's the point and that you shouldn't have to care, but I'm still curious.
Do you have any more details on what's actually running here?
Not that this is bad per se though.
Edit: They even pretty much copied the icons for browser, address book and dictionary.
Elementary might have something going on for them but originality is not it apparently.
The problem arises when you can trace so many ideas back to a single source that it makes it look like you're just making a cheap copy, which sadly is what is happening with elementary.
""" a better ISV story """
""" Channel issues """
This is the first step towards actually making inroads. That right there is BS and its the language of the corporate sales cycle. Linux will need advocates who are fluent in the language of the corporation to make the partnerships and sales deals that will be required to get Linux onto the corporate desktop. We are not there yet where we can walk into a sales room and compete on a corporate sales cycle with the professional sales guys who pitch solutions based on Microsoft, but at that point that's the battle that needs to be fought.
It isn't cable of running a local dev server or compiling code, which are things I still need from my computer (I'm hoping for a chrome os pro supporting guest vms) but everything else is so easy I see the os gaining serious ground and think it is silly to leave it out of a discussion of desktop linux.
Persoinally, I'm not so sure Linux needs to aim to be used by everyone under the sun... People have different needs and pluralism in operating systems (yes, including MS) can be a good thing....
Android is not suddenly going to be a desktop thing. It'll happen over time as Android devices become more and more important than desktops. An Android PC will serve to extend that Eco-system on the same level as iOS & OSX (but better since it's still fun).
> It has openness and it has the big support of Google.
Google did not open up Android 3.0 when it did not suit them to do so (commercially). I've never heard of Redhat or Canonical doing something similar with one of their Linux releases.
I'm not entirely comfortable with Google either, but I'm far less comfortable with Apple & M$.
Although if someone took Android and modified the application framework, windowing system, etc. to be a traditional WIMP desktop OS, I think that would definitely interest me.
No, I moved on because I realized that I don't like the desktop paradigm. The best thing I ever did to my computer was changing to tiling windowing and a keyboard oriented user experience. It's definitely not for everyone, but I don't miss the desktop.
Semantics. I prefer to call it a personal unix workstation :) I have a similar software setup on my laptop, which you can't call a "desktop". We can call it a portable unix workstation then.
According to the standard terminology used by most writers in the computer industry last 5 years, yes you can (when applied to software).
In contrast, if you say, "After using a desktop linux box for about 15 years, I have moved on," most readers are going to be misled into concluding that you've switched to OS X or Windows. (We only got what you actually meant because we were paying sufficient attention and because you went on to talk of tiling windowing and a keyboard oriented user experience, which are not really an option on OS X and Windows.)
P.S. In all the discussion about tablet computer, you really never noticed any of the thousands of authors using "desktop OS" to refer to the more-traditional kind of OS that runs on both desktop computers and laptop computers?
And you never noticed authors using "desktop" to distinguish a user-facing computer from a server?
Seriously, I've been reading about dreams of linux on the desktop for 20 years. Give up. Move on.
What matters to me is that there's nothing I have on my OS X laptop at work that I miss when I use my Linux laptop at home, but I can't say the same in reverse.
I use Linux a lot. But my home laptop has been OSX for the last 6 years, and it will stay OSX for the foreseeable future. Maybe it will never see mass adoption. Maybe it will. For my part it doesn't matter. What matters to me is that there's nothing I have on my Linux workstation that I miss when I use my OSX laptop at home, but I can't say the same in reverse.
Have you noticed that Linux is taking over the mobile market?
If you miss your N900, you can check what is happening with the Open Pandora device, and use your phone to theter your internet connection to it :)
In order to save the desktop we must kill the desktop.
Stop the madness and get on with life: apt-get install ratpoison
The desktop metaphor is the worst kind of skeuomorphic design. Why should I waste precious screen real estate to show a background image?
The only thing tiling window managers need for better adoption is to ship with better defaults and a reference card for keyboard shortcuts.
That works for hackers but for non-techies that aren't willing to put any time and effort into re-learning the basics.
What I hope to see is more tiling windowing features on mainstream user interfaces for the intermediate users.
And those phones will run Linux with the Android UI and drivers.
I believe tiling window managers would be fantastic productivity boosters for enterprise use. Less distractions for the worker bees.
You seem to assume Android is just another Linux distro with a different windows manager? If so, you are mistaken. An Android system is very different from a Linux distro - they basically only share the kernels and it is quite wrong to classify Android as "Linux" because it creates confusion.
Right now they mainly just share the kernel. If the GNU tool chain is included and you can compile and run linux programs, what is the difference besides some architectural changes?
That's a very big "if". It is not the case currently so I would not put them under the same umbrella.
- Many applications are designed with a floating environment in mind, and as a result simply translate very poorly to a tiling environment.
- If I'm working with more than three or four windows (which is common) I've basically got my active window and a stack of non-active, tiny, non-useable windows. In other words, it becomes difficult to combine even a medium-size set of windows with a usage model wherein I wish to interact with 2-3 medium size windows simultaneously.
- Often in a floating environment, I can overlap windows, maximizing utilized screen space by covering up unimportant parts of a certain window.
There are basically two reasons I keep using a tiling WM. First, it does pretty well with term emulators. Second, the existence of tabbed term emulators. The ability to spawn and destroy tabs on a whim without changing the screen layout really helps with my thought flow. If I was really a purist I'd be using GNU screen for that, but at that point what am I doing in X11 anyway?
This is my biggest issue as well. I would like a way to hide these windows and take them out of my next/previous stack. Out of all the tiling WMs out there I feel one has to have this feature but I've been so happy with ratpoison otherwise I haven't looked.
Perhaps, to stay in keeping with the ability to do everything from the keyboard, treat the "taskbar" like a second stack, only reduced to names instead of shrunken windows.
I never liked tiling window managers using the master-slave paradigm (like xmonad does) precisely because of the problems you both described.
http://i3wm.org/screenshots/ (check the screencast and skip to 2:55 for examples of the different layouts)